Pseudo-continuum modeling of wall-cooled packed-bed reactors with d t / d p ≤ 8 remains constrained by the accurate description of heat transfer, which arises from the coupling between bed-scale structural heterogeneity and fluid dynamics. Models commonly represent pseudo-local transport through effective transport parameters estimated by fitting model predictions to experimental or simulated data, limiting transferability. This work develops a PRS-informed pseudo-continuum heat-transfer model for an industrial-scale packed-bed reactor with d t / d p = 3.048 over the range R e p = 750 –1400. A representative unit cell is used as a reduced particle-resolved computational domain to characterize the local structural, fluid-dynamic, and heat-transfer behavior of the packed bed and to derive transport closures for the reactor-scale model. Axial and radial heat-transport descriptors are calculated directly from particle-resolved simulations (PRS) within this domain and transferred as internal reactor-scale closures without regression against reactor-scale temperature measurements. Under non-reactive conditions, the resulting formulation captures the averaged unit-cell thermal response and the principal axial and radial temperature trends measured in the industrial-scale reference system. Future work will expand the PRS database across reactor configurations and operating conditions to train and independently validate machine-learning surrogates for position-dependent transport closures.
No takes yet. Share an insight, caveat, or question.
Hernández-Aguirre et al. (2026) studied this question.
Synapse has enriched 5 related papers. Consider them for comparative context: